Quantitative Research

Information on statistical software

Proprietary and in-house statistical software for UNSW staff and students (access may be restricted)

SPSS

Study design

One of the first steps for any study is a definition of study aims. Generally, researchers are often interested in factors that are affecting or predicting an outcome variable (e.g. What are the risk factors for malaria infection?). For a quantitative study, study aims may need to be refined for the data that is TO BE collected (e.g. definition of the risk factors, measurements of malaria infection). The method of data collection should always be adequately described at an early stage of the study (or even before a study is being conducted). At this stage, one should consult with a statistician regarding definition of the data to be collected. For example:

• Are they categorical or continuous variables and how are you measuring and recording these variables?
• What is the expected distribution of the data (this may be obtained from literature or from experience)?

There are also study design issues. For example:

• Depending on the method of data collection, what study type should you use?
• What is the estimated sample size needed for the study?
• What are other factors that you need to control or account for in your study (again, this needs to be obtained from literature or from experience)?

Sample size issues

Study design in epidemiology

Design of Randomised Control Trials (RCTs)

Basic introduction: Chan Y. H. (2003) Randomised Controlled Trials (RCTs) - Essentials. Singapore Medical Journal 44(2): 060-063

Resources for common statistical procedures

Data entry and basic statistics

Obtaining some basic statistics (e.g. counts, means, standard deviations, box-plots, histograms) on your data is one of the first steps that is ESSENTIAL to ANY data analysis. It often gives you indications of errors in your data (e.g. implausible outliers, coding errors). It should also give you an indication of the following:

• Do you have a reasonably balanced data structure (e.g. are there too few data points in a particular factor level?)
• Do you need to perform data editing (e.g. correction of data errors, coding of missing values, grouping of data)?
• Do you need to perform transformation for skewed data?

At this stage, you might also check for potential correlations in your independent (or explanatory) variables so as to give you an idea if you might have problems with co-linearity when you are analysing and interpreting your data. Finally, your initial data exploration would give you an indication of the appropriate type of analysis to perform. You would probably need to do further basic statistics after performing an initial analysis to do further checks on your data.

Basic introduction: Chan Y. H. (2003) Biostatistics 101. Data presentation. Singapore Medical Journal 44(6): 280-285

Analysis of contingency tables

Step-by-step guide for SPSS by Prof Guang-Hwa Chang (Youngstown State University)
Relative risk versus Odds ratio

2-sample t-test

ANOVA

Parametric versus non-parametric tests

Basic introduction and step-by-step guide for SPSS: Chan Y. H. (2003) Biostatistics 102: Linear regression analysis. Singapore Medical Journal 44(8): 391-396

Linear regression

Basic introduction and step-by-step guide for SPSS: Chan Y. H. (2004) Biostatistics 201: Linear regression analysis. Singapore Medical Journal 45(2): 55-61
Technical notes (esp. for SPSS): Prof G. David Garson, NC State University
Pointers and FAQs: Steve Simon PhD, Children's Mercy Hospitals and Clinics

Logistic regression

Basic introduction and step-by-step guide for SPSS: Chan Y. H. (2004) Biostatistics 202: Logistic regression analysis. Singapore Medical Journal 45(4): 149-153
Pointers and FAQs: Steve Simon PhD, Children's Mercy Hospitals and Clinics

Multivariate analysis

Multi-level modelling

Repeated measures with missing values

Basic introduction and step-by-step guide for SPSS: Chan Y. H. (2004) Biostatistics 301A: Repeated measures analysis (mixed models). Singapore Medical Journal 45(10): 456-460

Survival analysis

Basic introduction and step-by-step guide for SPSS: Chan Y. H. (2004) Biostatistics 203: Survival analysis. Singapore Medical Journal 45(6): 249-256

Kaplan-Meier analysis

Technical notes (esp. for SPSS) from Prof G. David Garson, NC State University

Cox analysis

Factor analysis

Glossary of statistical terms

Statistics Glossary @ University of Glasgow